Today, HPE paid $650 to acquire Simplivity. HPE paid this sizeable amount mainly to acquire OmniStack, Simplivity’s data virtualization software platform. OmniStack’s central value proposition is OPEX reduction through the elimination of traditional infrastructure management tasks. In short, Simplivity enables customers to manage their infrastructure at the virtual machine (VM) – level. Please note that there will be a separate EMA Impact Brief covering this topic in much more detail.

Strategic Analysis
Today, HPE already offers its HPE HyperConverged Operating Environment 2.0 and numerous Hyper Converged hardware solutions. In addition, HPE just launched its “composable infrastructure” offering.

Then why spent $650m on Simplivity’s data virtualization layer technology? Because HPE’s converged products and composable infrastructure offering still include too many “moving parts” that have to be managed (WAN acceleration, hypervisor, file storage, block storage, data protection, cloud gateway, deduping and SSD and spinning disk arrays). This still provides more than enough ammunition for CFOs to raise an eyebrow and demand the move to Amazon Web Services or Microsoft Azure. And that exactly is why Simplivity was worth the $650m. Each workload moving to the public cloud is unlikely to ever return to HPE hardware or management tools. So by arming enterprise IT with a strong OPEX argument against AWS and Azure, OmniStack will be the dam that prevents even more application workloads from moving out of the data center for good.

Highlights of the IP that Cost $650m
• All operations management happens centrally and at the VM-level. Customers can leverage their existing management tools such as vSphere, vRealize or another platform able to manage VMware ESXi resource pools.
• VMs across any number of worldwide data centers are stored and managed within one central data virtualization layer. This means:
o They can be moved freely between Simplivity appliances, even across geographical locations.
o Deduplication (WAN optimization) enables almost instant backups, clones and rapid site replication.
o Clones and backups can be moved very fast, as only the delta is stored.
o Increased density, as only deltas between VMs are stored (also positive effect on CPU and RAM usage).
• VM-placement and data protection policies are centrally applied for all VMs across the entire enterprise.
• OmniStack enables linear scalability, as the data virtualization layer can be scaled across additional infrastructure resources across locations.

In Short: Simplivity’s Data Virtualization layer can become a great asset to HPE, if the company manages to leverage this technology across its vast portfolio.

 

Update (March 10, 2017): Now that the HPE acquisition of Simplivity has closed and the integration of Simplivity staff into the HPE organization has started, HPE answered some key questions regarding product roadmap and GTM.

“HPE Hyperconverged Systems used to loose due to a lack of data services, such as deduce, compression and data protection. Simplivity fills in this gap.” says one HPE executive. Of course, HPE also lost and still looses because of a lack of VM-centric management across server, rack, data center and cloud boundaries.

 

Here’s a quick summary of how HPE plans to leverage Simplivity OmnitStack Technology:

Priority 1: 

  • Simplivity “OmniStack On ProLiant DL380” ships as of February, but without UI/UX integration.
  • VM mobility between 3PAR storage and OmniStack.

Priority 2: 

  • Integration of Simplivity OmniStack with Hyper Converged Operating Environment v2 / OneView UX (2H, 2017). But only for the HPE HC380.
  • Support for Microsoft HyperV: HPE is aware of the importance to support more hypervisors than just VMware vSphere (timing unclear).
  • Integration with OneView operations management platform and Synergy Composer: HPE is aware of the criticality of these integration points
  • Integration with the rest of the HPE ProLiant server portfolio and with the new HPE Synergy composable platform.

Priority 3:

  • Support for Amazon Web Services and Microsoft: While HPE and Simplivity executive keep stressing the importance of keeping their hybrid infrastructure management systems “open”, the answers that were given regarding public cloud integration did not show a high level of enthusiasm for integration with the most popular public clouds.
  • Azure: While there seemed to be a close relationship between Simplivity and Microsoft, it is unclear today when Azure resources can finally be used to complement the HC380 or 250.
  • AWS: Nobody from HPE or Simplivity wants to comment on the priority of Amazon support, at least for specific scenarios, such as disaster recovery or high availability.

Conclusion: While I think that it is an OK start to focus on ProLiant, 3PAR and OneView UX integration, it is concerning to observe how HPE continuous to advertise an “open strategy”, without clearly committing to the support of public clouds. The lack of excitement regarding integration with Amazon, Azure and other public cloud offerings and the continued focus on mentioning mostly the downsides of the clouds does not boat well for a truly open strategy. In contrast, Cisco’s new HyperFlex strategy acknowledges  one of the old truths of enterprise IT (a great mentor of mine thought me this very early on in my career): “If you do not let your customers look over the fence to see what the competition is up to, these customers will become even more curious and eventually dump you.” So HPE advertises that you can save your money instead of having to buy Zerto for data protection. However, if I spend money on Zerto, I get full support of any bare metal, virtual, private or public cloud, without any lock-in. My hope is for HPE to recognize that lock-in is no longer viable today (there’s great EMA research data available to proof this fact) and that the “race up the stack” is on. Today, the vendor with the most flexible, open and well integrated management tools wins, while everyone who is unenthusiastic about managing “the other guy’s” environments can no longer be successful.